| Date: | July 26, 2004 / year-entry #288 |
| Tags: | history |
| Orig Link: | https://blogs.msdn.microsoft.com/oldnewthing/20040726-00/?p=38343 |
| Comments: | 20 |
| Summary: | Back in the days of Windows 95, when Plug and Play was in its infancy, one of the things the Plug and Play team did was push the PCI specification to an absurd extreme. They took a computer and put it at one end of a hallway. They then built a chain of PCI bridge cards... |
Back in the days of Windows 95, when Plug and Play was in its infancy, one of the things the Plug and Play team did was push the PCI specification to an absurd extreme. They took a computer and put it at one end of a hallway. They then built a chain of PCI bridge cards that ran down the hallway, and at the end of the chain, plugged in a video card. And then they turned it on. Amazingly, it actually worked. The machine booted and used a video card twenty feet away. (I'm guessing at the distance. It was a long time ago.) It took two people to operate this computer, one to move the mouse and type, and another to watch the monitor at the other end and report where the pointer was and what was happening on the screen. And the latency was insane. But it did work and thereby validated the original design. Other Plug and Play trivia: The phrase "Plug and Play" had already been trademarked at the time, and Microsoft had to obtain the rights to the phrase from the original owners. |
Comments (20)
Comments are closed. |
Is there a photo?
"It took two people to operate this computer…"
Why didn’t the bozo’s just take the chain of bridge cards out half way and then double back so the monitor and keyboard were next to each other?
Sounds like you need very long mouse and keyboard cable or a mirror and telescope.
Gee Jeff, they weren’t going to use it or anything.
Don’t forget about the time when Tom Adams created the spider-web of 255 USB devices all hooked up to the same PC to make sure that those limits worked too.
Let me guess, you were reminded of this feat by watching the Celtic/Chelsea match?
Maybe Microsoft was better at doing trademark searches back then. More recently, they seemed to have just started using the phrase "Caller ID for Email" even though it was already trademarked and in use on a product from a Texas company. Now that Microsoft’s CID for Email merged with SPF, they’ve redubbed it Sender ID. Maybe that was part of the settlement.
In the late 80s, I worked at a company that made storage boxes that hung off minicomputers. As a proud custodian of a VAX workstation, I had the privilege of running some of the prototypes. With one dual SCSI drive cabinet, we kept having problems. Sometimes it would see the first disk, sometimes neither. After ruling out configuration problems, we unscrewed the cabinet and looked inside. There were miles and miles of ribbon cable inside the box, enough to challenge the cable length limit in the SCSI spec. The short cable that connected the external box to the VAXstation pushed it beyond the limits required for timing and termination. We convinced the engineers to remove several unnecessary feet from inside the box.
Clearly they were setting up the wrong tests, because PCI plug and play worked only about 75% of the time in my experience. Blame it on who you will but in practice it was not plug and play for all users.
Adrian,
I had somewhat similar experience as well. Back in 1999 an ex-coworker of mine called me up and complained about his flaky computer. He’s a 3d artist and his computer knowledge lies between "MS sucks" and "wipe it out and reinstall the OS". So, he’d reinstall the OS every Friday (I’m not kidding) just to find out that at some point of time in the middle of next week his 40-hrs render sequence would hang the machine, ruining the whole progress.
It took several hours to inspect the configuration and run some tests. I couldn’t reproduce the hang and since sitting 40 hrs in front of the computer wasn’t among of my plans, I opened the box. It was a nicely configured machine made by one of the manufacturers whose ads are still all over the media, and whose name shall remain undisclosed. I found that a SCSI terminator was not fully plugged in. I honestly think the disk shouldn’t have worked at all, but it actually did, for some time at least. Maybe, contact was weak, but it worked until the inside of the box heated up enough to bend the plug and break the contact. Funny thing, the manufacturer actually left a spare terminator inside of the box. Talking about happy customers and reliability.
During the Windows 2000 project, we did something very similar. To test Plug and Play and to test the Driver Verifier, we created the "USB Cart of Death". We started with a 2 level cart similar to what you’d see in a library.
About ten 8port hubs were wired together, and then every port was filled with some different type of USB device. A USB steering wheel adorned the back of the cart, and a USB radio provided the antenna. Two cameras were one the front. All power went to a USB UPS. The entire cart, completely mobile, came down to two cables (power, USB). The final USB cable was plugged into a USB PCMCIA card.
We’d plug in the card into a laptop, watch the OS start up the 50 or so devices on it, and then (before or after it finished) we’d uncerimoneously yank the PCMCIA card. If a blue screen occured or the Driver Verifier detected a bug, the appropriate developer would be asked to look at the machine. In the meantime, the card would be wheeled to the next laptop, in hopes of finding a different bug.
7/26/2004 1:07 PM Adrian Oney:
> During the Windows 2000 project, we did
> something very similar.
I guess that’s why Windows 2000 didn’t blue-screen when a USB floppy drive was disconnected. Can you persuade the Windows XP project to duplicate your test?
(Hmm… PCMCIA USB cards are mostly USB2, right? The test ALSO needs duplication on built-in USB 1.1 ports. There’s a ton of other tests of things that worked or mostly worked in W2000 that need to be duplicated in XP too. Hmm, PCMCIA, have you heard about PCMCIA SCSI cards, and if so then do you still have any ears remaining?)
"I found that a SCSI terminator was not fully plugged in. I honestly think the disk shouldn’t have worked at all, but it actually did, for some time at least."
In my experience, SCSI drives will often work without proper termination… at least a little bit, occasionally, for some time, etc. Very nasty problem to track down if you forget to check the terminators.
Termination has surely got to be the first thing to check if there’s trouble with SCSI devices. Some say SCSI buses need three terminators – one at each end, and the termination of a chicken over the middle.
A wize man once told me that 90% of scsi problems are caused by one of 3 things. Cabling, IDs and termination.
Okay, this one’s supremely geeky, and forwarded to me from across the office. I think I know my audience though, so anyone with a yen to do so may read the tale of the twenty foot long computer, with diversions…
So my thoughts of building a PC completely wall mountable (each individual card mounted on its own piece of plastic then somehow connected to the motherboard) is actually possible given the proper parts? Neat.
Wow, i thought i would be the only one who wanted a wall mounted computer. I think it would be really cool to use shadow boxes/frames and try to market the idea